> Even with 10k codes (2 days) you get dozens of duplicates, so I'd say it's potentially even more common. Yeah I screwed up here, at the end: > Sqrt of a mil

> Even with 10k codes (2 days) you get dozens of duplicates, so I'd say it's potentially even more common. Yeah I screwed up here, at the end: > Sqrt of a million is a thousand, and if they're every 30 seconds, that's ~8 hours or so. So you probably get a duplicate or 2 every day. You'd get 1 or 2 every day (approx.) if you're only, at any point, looking at the collection of codes generated in the past 8 hours. But of course that wasn't the question, and the odds go way up once the period we're looking at goes larger and larger over time. I shouldn't have tried to go beyond "ballpark if you wait 8 hours you have a coinflip of having at least one duplicate", anything more than that requires different math.